Methods for joint modelling of multiple longitudinal and survival outcomes
Lead Research Organisation:
University of Cambridge
Abstract
When we want to predict the risk of a person developing a disease, or experiencing some other kind of medical event, we build a risk prediction model. The risk prediction model is used to inform someone about their future risk of disease and/or to make treatment decisions about the most appropriate way to manage their disease risk.
Usually a risk prediction model uses recent measurements of factors that are known to influence disease risk. We are exploring whether previous measurements of these risk factors, which are commonly stored in electronic medical records, can give additional information and help us to estimate the disease risk more precisely.
In this research programme we are exploring statistical methods which allow past measurements to be used in risk prediction, and developing statistical methods to determine whether a personalised screening schedule could help identify those at high disease risk.
One application of these methods is in cardiovascular risk prediction. This work will have important implications in the management of the risk of cardiovascular disease by improving the accuracy of cardiovascular risk prediction and developing a more individualised approach to the scheduling of cardiovascular risk assessments.
We will explore the potential of using data from electronic health records, such as records from GP practices, in this research. We anticipate this type of data will bring many challenges because the data were not originally collected for research purposes, and are considered “big data” because of the number of individuals involved.
Usually a risk prediction model uses recent measurements of factors that are known to influence disease risk. We are exploring whether previous measurements of these risk factors, which are commonly stored in electronic medical records, can give additional information and help us to estimate the disease risk more precisely.
In this research programme we are exploring statistical methods which allow past measurements to be used in risk prediction, and developing statistical methods to determine whether a personalised screening schedule could help identify those at high disease risk.
One application of these methods is in cardiovascular risk prediction. This work will have important implications in the management of the risk of cardiovascular disease by improving the accuracy of cardiovascular risk prediction and developing a more individualised approach to the scheduling of cardiovascular risk assessments.
We will explore the potential of using data from electronic health records, such as records from GP practices, in this research. We anticipate this type of data will bring many challenges because the data were not originally collected for research purposes, and are considered “big data” because of the number of individuals involved.
Technical Summary
In many clinical applications we are interested in statistically modelling more than one outcome. These might be multiple longitudinal outcomes, such as the risk factors blood pressure, total cholesterol and high-density lipoprotein cholesterol for cardiovascular disease (CVD), or they may include a time-to-event, such as the first CVD event. The aim of this research programme is to develop and explore statistical methodology for the analysis of multiple outcomes. One area of focus will be dynamic risk prediction, where an individual’s prediction for the risk of an event occurring is updated in response to new measurements made of time-varying risk predictors. Methods for modelling repeated measurements of multiple risk predictors and a time-to-event outcome include joint modelling, where all outcomes are modelled simultaneously, and landmarking, where past repeated measurements are modelled separately from future times-to-events in a two-stage approach.
A clinical application which motivates this work is cardiovascular risk prediction. We are using data from electronic health records to develop a risk prediction tool for cardiovascular disease and to provide recommendations for the optimal scheduling of cardiovascular risk assessments. By projecting the risk of cardiovascular disease into the future we will aim to determine whether a person will likely be at high risk of cardiovascular disease in the future, and thereby to obtain a personalised recommendation for the time to schedule the next risk assessment. Using data from electronic health records for clinical research can be challenging because this data has not been collected for research purposes. We will explore the impact of informative observation of risk factor measurements because these measurements may have been taken in response to the state of health of the individual. There will also be computational challenges in the application of complex statistical methods to such big data. Additional questions we will explore in this context include the improvement in predictive accuracy that could be achieved by including other aspects of the risk factor trajectory in a prediction model, such as the slope of the longitudinal trajectory, or the variability in the repeated measurements, and the use of decision theory in obtaining the personalised recommendations. This work will have important implications in the management of the risk of cardiovascular disease.
Another application which motivates this programme of research is to understand the relationship between lung function and survival in cystic fibrosis. We are using joint modelling to assess whether the effect of sex on survival in cystic fibrosis patients is mediated by lung function, and will further explore the role of body mass index.
A clinical application which motivates this work is cardiovascular risk prediction. We are using data from electronic health records to develop a risk prediction tool for cardiovascular disease and to provide recommendations for the optimal scheduling of cardiovascular risk assessments. By projecting the risk of cardiovascular disease into the future we will aim to determine whether a person will likely be at high risk of cardiovascular disease in the future, and thereby to obtain a personalised recommendation for the time to schedule the next risk assessment. Using data from electronic health records for clinical research can be challenging because this data has not been collected for research purposes. We will explore the impact of informative observation of risk factor measurements because these measurements may have been taken in response to the state of health of the individual. There will also be computational challenges in the application of complex statistical methods to such big data. Additional questions we will explore in this context include the improvement in predictive accuracy that could be achieved by including other aspects of the risk factor trajectory in a prediction model, such as the slope of the longitudinal trajectory, or the variability in the repeated measurements, and the use of decision theory in obtaining the personalised recommendations. This work will have important implications in the management of the risk of cardiovascular disease.
Another application which motivates this programme of research is to understand the relationship between lung function and survival in cystic fibrosis. We are using joint modelling to assess whether the effect of sex on survival in cystic fibrosis patients is mediated by lung function, and will further explore the role of body mass index.
Organisations
- University of Cambridge, United Kingdom (Lead Research Organisation)
- Hospital Sirio Libanes, Sao Paulo (Collaboration)
- Australian National University (ANU) (Collaboration)
- University College London, United Kingdom (Collaboration)
- University of California, Berkeley (Collaboration)
- University of Bristol, United Kingdom (Collaboration)
- University of Mississippi (Collaboration)
- Imperial College London, United Kingdom (Collaboration)
- London Sch of Hygiene and Trop Medicine, United Kingdom (Collaboration)
- Newcastle University, United Kingdom (Collaboration)
- The Cochrane Collaboration (Collaboration)
- Medical Research Council (Collaboration)
- University of Hamburg, Germany (Collaboration)
- University of Cambridge (Collaboration)
- University of Liverpool, United Kingdom (Collaboration)
- University of Helsinki, Finland (Collaboration)
- Lancaster University (Collaboration)
- University of Leicester, United Kingdom (Collaboration)
People |
ORCID iD |
Jessica Barrett (Principal Investigator) |
Publications

Barrett J
(2017)
Disease Modelling and Public Health, Part A

Barrett JK
(2019)
Estimating the association between blood pressure variability and cardiovascular disease: An application using the ARIC Study.
in Statistics in medicine


Cassell A
(2018)
The epidemiology of multimorbidity in primary care: a retrospective cohort study.
in The British journal of general practice : the journal of the Royal College of General Practitioners

Gasparini A
(2020)
Mixed-effects models for health care longitudinal data with an informative visiting process: A Monte Carlo simulation study.
in Statistica Neerlandica

Gasperoni F
(2020)
Evaluating the effect of healthcare providers on the clinical path of heart failure patients through a semi-Markov, multi-state model.
in BMC health services research

Grootes I
(2018)
Predicting risk of rupture and rupture-preventing reinterventions following endovascular abdominal aortic aneurysm repair.
in The British journal of surgery

Keogh R
(2018)
EPS5.01 Personalised time-updated predictions of short-term and long-term survival in CF using UK Registry data
in Journal of Cystic Fibrosis

Keogh RH
(2019)
Dynamic Prediction of Survival in Cystic Fibrosis: A Landmarking Analysis Using UK Patient Registry Data.
in Epidemiology (Cambridge, Mass.)

Law M
(2016)
Two new methods to fit models for network meta-analysis with random inconsistency effects.
in BMC medical research methodology
Description | Citation in paper on care planning for dementia patients |
Geographic Reach | Multiple continents/international |
Policy Influence Type | Citation in clinical reviews |
Description | Lecturer on short course, An introduction to R statistical software. |
Geographic Reach | Local/Municipal/Regional |
Policy Influence Type | Influenced training of practitioners or researchers |
Description | ANCA-associated vasculitis |
Organisation | University of Cambridge |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Statistical analysis to develop prediction models for relapse and infection in patients with ANCA-associated vasculitis. We used data from patients registered at an Addenbrooke's hospital clinic. |
Collaborator Contribution | Data acquisition, clinical expertise, writing the manuscript. |
Impact | A manuscript has been submitted to the Annals of Rheumatic Diseases. |
Start Year | 2019 |
Description | Blood pressure variability |
Organisation | University of Bristol |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Intellectual input, data analysis, writing |
Collaborator Contribution | Intellectual input |
Impact | One paper published in Statistics in Medicine |
Start Year | 2016 |
Description | Blood pressure variability |
Organisation | University of Mississippi |
Department | Department of Data Science |
Country | United States |
Sector | Academic/University |
PI Contribution | Intellectual input, data analysis, writing |
Collaborator Contribution | Intellectual input |
Impact | One paper published in Statistics in Medicine |
Start Year | 2016 |
Description | Breast cancer systematic reviews |
Organisation | Hospital Sirio Libanes, Sao Paulo |
Country | Brazil |
Sector | Hospitals |
PI Contribution | Intellectual input, writing and data analysis. |
Collaborator Contribution | Intellectual input, literature searches, data extraction and analyses. |
Impact | One Cochrane review has been published. A second Cochrane review is under review. A third paper has been published in Reports of Practical Oncology and Radiotherapy. |
Start Year | 2012 |
Description | Breast cancer systematic reviews |
Organisation | The Cochrane Collaboration |
Department | Brazilian Cochrane Centre (BCC) |
Country | Brazil |
Sector | Charity/Non Profit |
PI Contribution | Intellectual input, writing and data analysis. |
Collaborator Contribution | Intellectual input, literature searches, data extraction and analyses. |
Impact | One Cochrane review has been published. A second Cochrane review is under review. A third paper has been published in Reports of Practical Oncology and Radiotherapy. |
Start Year | 2012 |
Description | Cardiovascular disease risk prediction using electronic health records |
Organisation | Australian National University (ANU) |
Department | National Centre for Epidemiology and Population Health |
Country | Australia |
Sector | Academic/University |
PI Contribution | Intellectual input and statistical analysis |
Collaborator Contribution | Intellectual input and data provision |
Impact | One paper published by the American Journal of Epidemiology, one paper in preparation. |
Start Year | 2013 |
Description | Cardiovascular disease risk prediction using electronic health records |
Organisation | University College London |
Department | Department of Chemistry |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Intellectual input and statistical analysis |
Collaborator Contribution | Intellectual input and data provision |
Impact | One paper published by the American Journal of Epidemiology, one paper in preparation. |
Start Year | 2013 |
Description | Cystic fibrosis |
Organisation | Lancaster University |
Department | Faculty of Health and Medicine |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Intellectual input, data analysis, writing. |
Collaborator Contribution | Intellectual input, writing |
Impact | Article published in JRSSB, DOI 10.1111/rssb.12060. A second paper is ready for submission. |
Start Year | 2011 |
Description | Cystic fibrosis |
Organisation | Newcastle University |
Department | School of Mathematics and Statistics |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Intellectual input, data analysis, writing. |
Collaborator Contribution | Intellectual input, writing |
Impact | Article published in JRSSB, DOI 10.1111/rssb.12060. A second paper is ready for submission. |
Start Year | 2011 |
Description | Cystic fibrosis |
Organisation | University of Liverpool |
Department | Institute of Psychology, Health and Society |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Intellectual input, data analysis, writing. |
Collaborator Contribution | Intellectual input, writing |
Impact | Article published in JRSSB, DOI 10.1111/rssb.12060. A second paper is ready for submission. |
Start Year | 2011 |
Description | Cystic fibrosis landmarking |
Organisation | London School of Hygiene and Tropical Medicine (LSHTM) |
Department | Medical Statistics |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Intellectual input and computing code |
Collaborator Contribution | Intellectual input, computing code and data analysis. |
Impact | One paper published in Epidemiology. |
Start Year | 2016 |
Description | Prediction of post-surgery rupture of abdominal aortic anuerysms |
Organisation | Imperial College London |
Department | Department of Surgery and Cancer |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Intellectual input, data analysis |
Collaborator Contribution | Intellectual input, data provision |
Impact | One paper published as a Health Technology Assessment report, one paper published in the British Journal of Surgery. |
Start Year | 2015 |
Description | Prediction of post-surgery rupture of abdominal aortic anuerysms |
Organisation | University of Hamburg |
Department | Department of Vascular Medicine |
Country | Germany |
Sector | Academic/University |
PI Contribution | Intellectual input, data analysis |
Collaborator Contribution | Intellectual input, data provision |
Impact | One paper published as a Health Technology Assessment report, one paper published in the British Journal of Surgery. |
Start Year | 2015 |
Description | Prediction of post-surgery rupture of abdominal aortic anuerysms |
Organisation | University of Helsinki |
Country | Finland |
Sector | Academic/University |
PI Contribution | Intellectual input, data analysis |
Collaborator Contribution | Intellectual input, data provision |
Impact | One paper published as a Health Technology Assessment report, one paper published in the British Journal of Surgery. |
Start Year | 2015 |
Description | Prediction of post-surgery rupture of abdominal aortic anuerysms |
Organisation | University of Leicester |
Department | Department of Health Sciences |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Intellectual input, data analysis |
Collaborator Contribution | Intellectual input, data provision |
Impact | One paper published as a Health Technology Assessment report, one paper published in the British Journal of Surgery. |
Start Year | 2015 |
Description | Splines for joint modelling |
Organisation | Medical Research Council (MRC) |
Country | United Kingdom |
Sector | Public |
PI Contribution | Intellectual input and writing computer code |
Collaborator Contribution | Intellectual input |
Impact | One paper published in Statistics in Medicine, and one paper published in Biometrics. |
Start Year | 2013 |
Description | Splines for joint modelling |
Organisation | University of California, Berkeley |
Department | Department of Integrative Biology |
Country | United States |
Sector | Academic/University |
PI Contribution | Intellectual input and writing computer code |
Collaborator Contribution | Intellectual input |
Impact | One paper published in Statistics in Medicine, and one paper published in Biometrics. |
Start Year | 2013 |
Description | Cambridge Science Festival 2019 |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Public/other audiences |
Results and Impact | Stand at Cambridge Science Festival 2019, presenting two interactive hands-on activities developed by researchers across the BSU; one explaining probability and risk, and the other on precision medicine. Reaching out to 500+ audience members over 1 day. |
Year(s) Of Engagement Activity | 2019 |
Description | Channel Network Conference talk |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Other audiences |
Results and Impact | A contributed talk at the Channel Network Conference, resulting in increased interest in my work and an invitiation to speak at a workshop. |
Year(s) Of Engagement Activity | 2019 |
Description | ISTAART webinar |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Other audiences |
Results and Impact | Webinar for the Design and Data Analytics group of the International Society to Advance Alzheimer's Research and Treatment (ISTAART). Over 40 were registered to attend, resulting in increased interest in the use of my methods and code in other applications. |
Year(s) Of Engagement Activity | 2020 |
Description | Introduction to 'R': free software for statistical analysis |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Industry/Business |
Results and Impact | About 30 professionals from different sectors attended the course. This is an introduction to R, which is a widely spread free software for analysing data. It is used by both researchers in academia and professionals in industry. The main outcome consists in learning the basics of R (i.e., syntax, algebra, functions, plots) and in learning how to retrieve information for specific questions about data analysis. |
Year(s) Of Engagement Activity | 2019,2020 |
URL | https://www.mrc-bsu.cam.ac.uk/training/short-courses/an-introduction-to-r/ |
Description | Invited talk at the 12th International Conference of the ERCIM WG on Computational and Methodological Statistics (CMStatistics 2019) |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | I was invited to give a talk by Dr. Maria Luisa Restaino on Survival Analysis. I presented my work on "Evaluating the effect of healthcare providers through semi-Markov multi-state model and nonparametric discrete frailty", which is now a submitted paper. The audience was composed of statisticians coming from different European countries. I was asked several questions at the end of the talk (about data collection and proposed model). I also discussed with the chair possible further developments of this work. |
Year(s) Of Engagement Activity | 2019 |
URL | http://cmstatistics.org/RegistrationsV2/CMStatistics2019/viewSubmission.php?in=1219&token=9r453rnor6... |
Description | Press coverage for work on epidemiology of multimorbidity in primary care |
Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Media (as a channel to the public) |
Results and Impact | Kirsty Rhodes co-author on paper: 'The epidemiology of multimorbidity in primary care: a retrospective cohort study' which received wide-spread national media coverage, including articles in The Telegraph, Express, Sun, Daily Mail and GP Online. https://www.telegraph.co.uk/news/2018/03/13/one-four-adults-have-multiple-health-problems-startling-data/ |
Year(s) Of Engagement Activity | 2018 |
URL | http://bjgp.org/content/early/2018/03/12/bjgp18X695465 |
Description | Seminar (Manchester) |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | Local |
Primary Audience | Other audiences |
Results and Impact | Seminar about cardiovascular risk prediction to an audience of around 40 from the University of Manchester, sparking questions and debate afterwards. |
Year(s) Of Engagement Activity | 2019 |